{"slug":"biaslens","name":"BiasLens","hackathon":"qhacks-2026","tagline":"Upload a trade log to see scores for overtrading, loss aversion and revenge trading, with explanations and coaching.","placement":{"status":"track-winner","awards":["National Bank of Canada Prize Track"],"track":null},"team":{"size":4,"members":[]},"problem":"Retail traders lose returns to cognitive biases they do not recognise in their own behaviour.","solution":"Computes behavioural features from a CSV of trades, scores three biases with a gradient-boosted model, explains each score with SHAP feature charts, flags the trades that triggered it and offers AI coaching.","demo_moment":"Team-described: upload trades and expand a bias card to see which behaviours drove it.","build_style":"software-only","tech":{"stack":["python","fastapi","xgboost","shap","scikit-learn","next.js","gemini","docker"],"ai_patterns":["ml-model","llm-api"]},"domains":["fintech"],"analysis":{"why_it_stood_out":[{"claim":"Won “National Bank of Canada Prize Track” among 77 submissions.","basis":"organizer-stated","source":1},{"claim":"Explains predictions with feature attributions rather than only flagging trades.","basis":"team-stated","source":0},{"claim":"Won a crowded sponsor challenge: many teams built bias detectors for the same brief.","basis":"curator-inference"}],"mechanisms":["cited-evidence-trust","visible-engineering-depth"],"judging_lens":["technical-depth","sponsor-api-use"],"lessons":["In a crowded sponsor brief, explainability is what separates near-identical detectors."],"confidence":"medium"},"links":{"devpost":"https://devpost.com/software/biaslens-2yd3g9","repo":"https://github.com/AlpinSchool/QHacks_2026","demo":"https://frontend-two-red-5qyy3cnl69.vercel.app/","video":null},"sources":[{"url":"https://devpost.com/software/biaslens-2yd3g9","kind":"devpost-project","retrieved_at":"2026-10-13"},{"url":"https://qhacks-2026.devpost.com/project-gallery","kind":"devpost-gallery","retrieved_at":"2026-10-13"}]}
